基于光谱特征的北京城区植被滞尘分布反演

发布时间：2018-05-21 00:37

本文选题：遥感 + 高光谱　；参考：《光谱学与光谱分析》2016年09期

【摘要】：植被叶片的滞尘量可以表征空气污染的程度,分析城市植被滞尘的空间特征对于制定更为有效的空气污染控制政策具有重要的现实意义。基于北京市主城区采集的大叶黄杨、国槐、毛白杨和山桃等四种典型绿化植被叶片的滞尘量、光谱反射率和叶面积等数据,比较四种植被叶片滞尘前后的光谱曲线,进行窄波段与卫星波段滞尘前后叶片光谱反射率比值与滞尘量的相关分析。然后,分别建立相关性最大的卫星波段反射率和NDVI与滞尘量之间的回归模型,选取拟合较好的模型反演北京城区植被的滞尘量分布,进而插值得到整个北京城区的尘埃分布。最后,根据高滞尘区域周围的土地覆盖和土地利用以及滞尘期间PM10浓度的空间分布对反演的的合理性进行检验。结果表明:在780~1 300nm波段,大叶黄杨、国槐、毛白杨和山桃四种植被的滞尘叶片反射率均明显低于干净叶片;窄波段反射率与滞尘量在520~650nm波段和1 390~1 600nm波段具有较高的相关性,相关系数的绝对值最高达到0.626;利用Landsat8的green波段和NDVI构建的滞尘反演模型,决定系数(R2)分别为0.446和0.465。NDVI模型反演的北京城区植被的滞尘量分布结果表明,北京城区滞尘含量呈现出北高南低,东高西低,中心城区高于郊区的空间分布格局。该研究通过高光谱和遥感影像数据反演滞尘量,可以为快速全面监测城市地区尘埃分布提供参考。
[Abstract]:The dedusting amount of vegetation leaves can represent the degree of air pollution. It is of great practical significance to analyze the spatial characteristics of urban vegetation dust retention for making more effective air pollution control policies. Based on the data collected from four typical greening vegetation in Beijing, such as Populus japonica, Sophora japonica, Populus tomentosa and Peach, the spectral reflectance and leaf area were compared before and after dust retention. The correlation between the ratio of leaf spectral reflectance and the amount of dust lag in narrow band and satellite band before and after dedusting was analyzed. Then, the regression models of satellite reflectance and NDVI and dust retention are established respectively, and the model with good fitting is selected to retrieve the distribution of dust retention of vegetation in Beijing urban area, and then the dust distribution of the whole Beijing urban area is obtained by interpolation. Finally, the rationality of the inversion is tested according to the spatial distribution of land cover and land use around the high dust retention area and the concentration of PM10 during dust retention. The results showed that the reflectance of dedusting-leaf leaves in 7801 300nm, Populus japonica, Sophora japonica, Populus tomentosa and Prunus amurensis was significantly lower than that of clean leaves, and the reflectivity of narrow band had a high correlation with dust retention in 520~650nm and 1390 600nm bands. The absolute value of correlation coefficient is 0.626, and the distribution of dust retention of vegetation in Beijing is obtained by using Landsat8's green band and NDVI's model of dedusting-dust inversion, which are 0.446 and 0.465.NDVI models respectively. The dust content of Beijing urban area is lower than that of north high and south, east high and west low, and the spatial distribution pattern of central city is higher than that of suburb. This study can provide a reference for rapid and comprehensive monitoring of dust distribution in urban areas by retrieving dust retention from hyperspectral and remote sensing image data.
【作者单位】：首都师范大学三维信息获取与应用教育部重点实验室;首都师范大学资源环境与地理信息系统北京市重点实验室;成都理工大学旅游与城乡规划学院;香港理工大学土地测量及地理资讯学系;
【基金】：国家青年科学基金项目(41201488) 国家基础测绘基金项目(2011A2011) 高等学校博士学科点专项科研基金项目(20111108110004)资助
【分类号】：X513;X173

【相似文献】